-
Synthesis of Rotaxanes and Pseudorotaxanes as
Intermediates towards Molecular Cages
by
Damian Franz Weske
A thesis submitted to the University of Birmingham for the
degree of
Doctor of Philosophy
School of Chemistry
University of Birmingham
January 2013
-
University of Birmingham Research Archive
e-theses repository This unpublished thesis/dissertation is
copyright of the author and/or third parties. The intellectual
property rights of the author or third parties in respect of this
work are as defined by The Copyright Designs and Patents Act 1988
or as modified by any successor legislation. Any use made of
information contained in this thesis/dissertation must be in
accordance with that legislation and must be properly acknowledged.
Further distribution or reproduction in any format is prohibited
without the permission of the copyright holder.
-
ii
Abstract
The work described in this thesis concerns the synthesis of
novel rotaxanes and
pseudorotaxanes as intermediates towards molecular cages. The
area of
rotaxanes and cage molecules was reviewed and selected examples
discussed in
detail.
Three main strategies to synthesise a fumaramide thread were
investigated,
however these were unsuccessful. By changing the templating
station and the
stopper groups, several novel azo-dicarboxylate threads were
synthesised, as well
as their [2]- and [3]rotaxanes. Unusual reductions of the
azo-dicarboxylate stations
during the clipping processes were observed and these were
studied in detail.
A couple of different stopper groups from those initially used,
provided several
novel double fumaramide threads and rotaxanes in high yields
Linking reactions of the two macrocycles of the newly
synthesised [3]rotaxanes to
form a molecular cage were investigated. However, reduction of
the macrocycles
was required, which led to deslipping of the macrocycles.
Therefore, new routes to
generate rotaxanes with bulkier stopper groups were
explored.
In the final section, novel pseudorotaxanes were synthesised and
detected via
1H NMR spectroscopy and mass spectrometry. By linking the two
macrocycles of a
[3]pseudorotaxane, a molecular cage was formed.
-
iii
Acknowledgments
First of all I would like to thank Professor Nigel S. Simpkins
for the opportunity to
undertake a PhD in his laboratory, for the interesting and
challenging project, as
well as for his guidance over the past three years.
Secondly, I would like to thank the School of Chemistry and the
University of
Birmingham for their financial support. I would also like to
thank the Analytical
Facility, specially to Graham Burns, Peter Ashton, Dr. Neil
Spencer, Dr. Louise
Male and Dr. Chi Tsang for their hard work.
I would like to thank the past and present members of the
Simpkins group. Special
thanks go to Fred, Bick, Ilias, Pete, Matt, Giorgio, Bhaven,
Shamim, Luke, Bene,
Lucia, Carlos, Andrea, Luca and Sebastien for all the good times
together.
I am grateful to Dr. Richard Grainger and Dr. Jim Tucker for
their advice during my
project and keeping always an open door for my questions.
I would also like to thank Mike, Tom, Dave, Huy, Bhaven and
Natalia for their help
at some point during my PhD.
A big thanks to Jenna and Chris for the nice days out and to
Mark and Amanda for
a great trip across Scotland and their overwhelming
hospitality.
I would like to express my deepest gratitude to Carina, Nora,
Katharina and my
mum, for their love and support during all my life.
-
iv
Last but not least, I would like to thank Natalia for making my
life incredibly
amazing, for all her support, love and humour that made my time
in Birmingham
the best of my life.
-
v
Table of Contents
1 Chapter 1: Introduction
....................................................................................
1
1.1 Supramolecular chemistry
........................................................................
1
1.2 Rotaxanes and catenanes
.......................................................................
2
1.2.1 Nomenclature of rotaxanes
.................................................................
5
1.2.2 Hydrogen bonding based rotaxanes
................................................... 7
1.2.3 Higher order rotaxanes
.....................................................................
11
1.2.4 Shuttling
............................................................................................
12
1.2.5 Rotaxanes used for sensing
.............................................................
16
1.2.6 Threaded structures with cross-linked macrocycles
......................... 17
1.3 Cage structures
......................................................................................
21
1.3.1 Metal organic cages
..........................................................................
22
1.3.2 Boronic ester based cages
...............................................................
23
1.3.3 Imine cages
......................................................................................
25
2 Chapter 2: Results and Discussion
...............................................................
27
2.1 Research aims
.......................................................................................
27
2.2 Attempted syntheses of the double fumaric thread 21
........................... 29 2.2.1 Synthetic route A towards
thread 21 ................................................. 30
2.2.2 Synthetic route B towards thread 21
................................................. 36 2.2.3
Synthetic route C towards thread 21
................................................. 41
2.3 Synthesis of azo rotaxanes
....................................................................
44
2.3.1 Synthesis of azo rotaxanes with a three carbon spacer unit
............. 47
2.3.2 Synthesis of azo rotaxanes with a four carbon spacer unit
............... 53
2.3.3 Synthesis of azo rotaxanes with a five carbon spacer unit
............... 57
2.3.4 Spontaneous reduction of azo systems
............................................ 65
2.4 Synthesis of fumaric rotaxanes
..............................................................
68
2.4.1 Synthesis of a fumaric [2]rotaxane with dibenzylamide
stopper .......... groups
...............................................................................................
68
2.4.2 Synthesis of a fumaric [3]rotaxane with a two carbon
spacer unit .... 71
2.4.3 Synthesis of a fumaric [3]rotaxane with a three carbon
spacer unit .. 74
-
vi
2.4.4 Synthesis of a fumaric [3]rotaxane with a three carbon
spacer unit ..... and a phenylbenzyl amide stopper
................................................... 77
2.5 Further reactions towards a molecular cage
.......................................... 80
2.5.1 Reduction of the tetraamide macrocycles
......................................... 81
2.5.2 Attempts to use alternative stopper groups
...................................... 88
2.6 Pseudorotaxanes
...................................................................................
98
3 Chapter 3: Conclusions and Future Work
................................................... 117
4 Chapter 4:
Experimental..............................................................................
121
4.1 Materials and methods
.........................................................................
121
4.2 Experimental for Section 2.2
................................................................
123
4.3 Experimental for Section 2.3
................................................................
133
4.4 Experimental for Section 2.4
................................................................
153
4.5 Experimental for Section 2.5
................................................................
168
4.6 Experimental for Section 2.6
................................................................
190
Appendix
............................................................................................................
201
References
.........................................................................................................
230
-
vii
Glossary
aq. aqueous
BtH benzotriazole
cat. catalyst, catalytic
conc. concentrated
CuAAC copper(I)-catalysed azide-alkyne cycloaddition
Da dalton
DBU 1,8-diazabicyclo[5.4.0]undec-7-ene
decomp. decomposition
DFT density functional theory
DMAP 4-(dimethylamino)pyridine
DMF N,N-dimethylformamide
DMSO dimethylsulfoxide
DNA deoxyribonucleic acid
EDC 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide
EI electron impact
eq. equivalent
-
viii
ES+ electrospray, positive ionisation mode
ES- electrospray, negative ionisation mode
HATU
1-[bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]
pyridinium 3-oxid hexafluorophosphate
HOAt 1-hydroxy-7-azabenzotriazole
HoBt hydroxybenzotriazole
HPLC high pressure liquid chromatography
IBCF isobutyl chloroformate
IR infrared spectroscopy
Lit. literature value
MALDI matrix assisted laser desorption ionisation
m.p. melting point
MS mass spectrometry
NBS N-bromosuccinimide
NMM N-methylmorpholine
NMR nuclear magnetic resonance
pet petroleum ether
ppm parts per million
-
ix
py pyridine
RT room temperature
Rt retention time
TLC thin layer chromatography
ToF time of flight
Å Angström (10-10 meters)
frequency
-
Chapter 1
Introduction
1
1 Chapter 1: Introduction
1.1 Supramolecular chemistry
In 1987 the Nobel prize in chemistry was awarded to Cram,
Pedersen and Lehn.
Jean-Marie Lehn is still one of the leading proponents in the
area of
supramolecular chemistry and defined it as “chemistry beyond the
molecule”.1 The
most important attribute of a supramolecular system is that
several components
(host and guest) are held together via reversible intermolecular
interactions. In
contrast to supramolecular systems, molecular systems are mainly
connected via
covalent bonds. Supramolecular chemists are able to synthesise
covalently
bonded molecular building blocks, which are held together by
intermolecular
forces to create functional architectures.2 These intermolecular
forces can be
electrostatic (ion-ion, ion-dipole and dipole-dipole), hydrogen
bonding, -
stacking, van der Waals, hydrophobic or solvatophobic
interactions. These
non-covalent interactions are generally much weaker than
covalent bonds. The
combination of several of these weak interactions generates a
powerful interlocked
system, and thus host molecules are able to selectively
recognise a specific guest.
As these interactions are reversible, host-guest products can
sit either in a local or
global thermodynamic energy minimum.
This type of lock and key principle can be compared to that used
by Nature itself,
e.g. in the reaction of a substrate with the active site of an
enzyme as described by
Emil Fischer in 1894.3 Base pairing in DNA is another example,
where two
-
Chapter 1
Introduction
2
antiparallel strands are held together by hydrogen bonds between
complementary
base pairs.4 The receptor sites on the host (lock) are
sterically and electronically
complementary to the guest (key). Therefore, the host is highly
selective for only
one substrate favouring one single product.
1.2 Rotaxanes and catenanes
Rotaxanes and catenanes are the two main classes of mechanically
interlocked
molecules.5 Mechanically interlocked molecules possess two or
more separate
covalent components. These components are connected via
mechanical bonds,
which are described as bonds that prevent dissociation. When one
of the covalent
bonds is cleaved, the components will dissociate. Catenanes are
formed of two or
more interlocked macrocycles (Figure 1.1)6, while rotaxanes are
made of at least
one macrocycle and one thread with two stoppers, which prevent
the dissociation
of the components (Figure 1.2).7,8
Figure 1.1: Structure of a [2]catenane.6
Catenanes are topological isomers and their rings cannot be
dissociated without
notionally breaking at least one covalent bond. In the other
hand, rotaxanes can
be separated via deformations of one of the components, e.g. a
macrocycle can
-
Chapter 1
Introduction
3
stretch and this might cause its slipping over the stoppers and
therefore
dissociation from the thread. In other words, rotaxanes are
threaded species
stabilised by steric interactions.
Figure 1.2: First [2]rotaxane synthesised by Harrison and
Harrison (left)9 and the graphical representation of a [2]rotaxane
(right).
The term rotaxane was first proposed by Schill and Zollkopf in
1967.10 The
nomenclature used is derived from the Latin where the word rota
means wheel
and the word axis means axle. In the same year, Harrison and
Harrison
synthesised the first [2]rotaxane with an overall yield of 6%
(Figure 1.2).9
Due to the fascinating structure of rotaxanes, the development
of straightforward
procedures to synthesise this kind of compounds has been
extensively explored,
leading to synthetic strategies which generally follow one of
three main routes.
These three methods involve threading followed by stoppering,
slipping or clipping
(Figure 1.3).11-19
In the threading method, a preformed macrocycle is first
coordinated onto a thread
via attractive interactions. Stopper groups are then attached to
the end of the
thread, which prevent dissociation of the newly formed
rotaxane.11 In the slipping
strategy, a pre-synthesised macrocycle is slipped over an
already stoppered
thread. The macrocycle and the stoppers have to be of the
correct size to allow
-
Chapter 1
Introduction
4
the formation of the rotaxane. The rotaxane is formed by
applying the right amount
of thermal energy, e.g. by lowering the temperature, the
macrocycle becomes
kinetically trapped, thus forming the rotaxane.19 The third
approach is the clipping
method, where a macrocycle is formed in the presence of a
thread.16 Usually an
attractive interaction between the ring forming components and
the thread is used
to initially entrap the macrocycle to be, followed by a
ring-closing reaction to form
the rotaxane.
Figure 1.3: The three main methods for rotaxane assembly.
In 2007, Leigh and co-workers developed a new strategy for the
synthesis of
rotaxanes called the active metal template method, where a metal
is used to
promote the formation of the thread and the coordination of the
preformed
macrocycle (Figure 1.4).20,21 This method was directly inspired
by the work of
Sauvage.20
-
Chapter 1
Introduction
5
Figure 1.4: Active-metal template method for rotaxane formation;
(i) template assembly and covalent bond forming catalysis, (ii)
subsequent demetalation.
The two stopper units (shown in green and orange) generate the
thread by the aid
of the metal (shown in pink). At the same time the metal
coordinates through the
cavity of the macrocycle (shown in black) via metal to ligand
interactions. In many
cases, once the rotaxane is formed the metal can bind to another
free macrocycle,
as this is more favoured and can promote the next threading
process.
1.2.1 Nomenclature of rotaxanes
Harrison and Harrison’s first rotaxane shows the simplest
rotaxane topology
(Figure 1.2). It is designated as a [2]rotaxane, where the
number in square
brackets indicates the number of components, two, that is one
macrocycle and
one thread.
Generally, an [n]rotaxane consists of one thread and (n – 1)
macrocycles or (n - 1)
threads and one macrocycle. Figure 1.5 shows some fundamental
rotaxane
topologies.
-
Chapter 1
Introduction
6
Figure 1.5: Some rotaxanes topologies.
[3]- and [4]rotaxanes are the corresponding higher analogues of
the [2]rotaxane,
with [3]rotaxanes being more common than [4]rotaxanes.21,22
The development of [2]rotaxanes with two separate sites or
stations has been an
important breakthrough for the synthesis of molecular switches,
motors and
machines.23 These two stations, that can be either identical or
different, can act as
binding sites for the macrocycle. When the two stations are
different, they can be
addressed selectively using an external input. Hence, chemical,
electrochemical,
or photochemical stimulation can induce changes in the binding
site by generating
repulsive or attractive supramolecular interactions and
therefore, the macrocycle
can be ‘shuttled’ between the two stations along the thread (see
Section 1.2.4).
This induced discrimination process between the two stations is
not possible when
the two sites are identical, however shuttling can still
occur.24
-
Chapter 1
Introduction
7
1.2.2 Hydrogen bonding based rotaxanes
Hydrogen bond interactions are known to assemble and stabilise
various
supramolecular complexes, such as catenanes and rotaxanes.25 The
first example
of a hydrogen bond templated catenane was reported by Hunter et
al. in 1992.26
High dilution for the macrocyclisation yielded a [2]catenane in
34% yield,
accompanied by the expected macrocycle in 51% yield. In 1995,
Vögtle et al.
reported the first example of an amide based [2]rotaxane via a
three component
threading procedure in a 11% yield.27 This threading process was
later improved
by employing a more rigid thread with the preparation of a
fumaric thread, which
yielded the desired [2]rotaxane in 18% yield.28 Leigh and
co-workers have also
described various examples of catenanes and rotaxanes templated
via hydrogen
bonding interactions. These rotaxanes are produced via a five
component clipping
reaction, as shown in Scheme 1.1.29 Reaction of isophthaloyl
dichloride and
p-xylylenediamine forms the intermediate amide I in a syn anti
conformation. In
solution, this intermediate can undergo either a dimerisation,
as described by
Hunter30 or a catenane formation process, as reported by
Leigh.31 However, in the
presence of a suitable template (e.g. peptide thread 1) a syn
syn conformation
(intermediate II) is favoured, as it is stabilised by
multicomponent bifurcated
hydrogen bonds. The terminal amine and the acid chloride are
then close enough
and a rapid cyclisation around the thread occurs (Scheme
1.1).29
-
Chapter 1
Introduction
8
Scheme 1.1: Rotaxane formation around a peptide thread.
Several oligopeptide threads bearing at least one non N-terminal
glycine residue
were also investigated by Leigh and co-workers.29 Figure 1.6
shows the crystal
structures for the obtained Gly-Gly 3, Gly-Sar 4 and Gly-L-Ala 5
[2]rotaxanes. By
comparing the yields of the resulting [2]rotaxanes with the
conformational
structures of their macrocycles, Leigh rationalised that there
is a relationship
-
Chapter 1
Introduction
9
between the final rotaxane and the intermediate II, which is key
for the formation of
the resulting rotaxane.
Gly-Gly 3 Gly Sar 4 Gly-L-Ala 5
Figure 1.6: Crystal structures of three different peptide based
[2]rotaxanes.29
The Gly-Gly rotaxane 3 has the least steric hindrance for the
assembly of the
pre-macrocycle around the thread, therefore the most favourable
hydrogen bond
interaction is realised in the chair conformation and 3 is
obtained in a 62% yield.
The Gly-Sar rotaxane 4 (60% yield) has an additional methyl
group which distorts
the hydrogen boding between the macrocycle and the thread and
consequently
the macrocycle has a half chair / twist boat like conformation.
The macrocycle of
the Gly-L-Ala rotaxane 5 has an unfavoured boat conformation and
this rotaxane
is obtained in only 45% yield.29
In 2001, Leigh and co-workers reported the first highly
efficient template-directed
synthesis of an amide based [2]rotaxane.32 The very rigid
fumaric thread 6 is able
to template the formation of the [2]rotaxane 7 via clipping
reaction of 4 equivalents
-
Chapter 1
Introduction
10
of p-xylylenediamine and isophthaloyl dichloride in a solvent
mixture of chloroform
and acetonitrile (9 : 1) (Scheme 1.2). Filtration and
spontaneous crystallisation
delivered the clean product in 97% yield.32
Scheme 1.2: Highly efficient [2]rotaxane formation.
It is believed that the high yield was due to the incorporation
of multiple
cooperative binding sites on a structurally rigid backbone,
facilitating the hydrogen
bond assisted synthesis of the rotaxane 7. The effective
templation of the
tetraamide macrocycle in a chair conformation around the rigid
fumaramide thread
6 maximises the hydrogen bond interactions between the two
substrates, making
this method one of the most accessible and efficient syntheses
described to
date.32
-
Chapter 1
Introduction
11
1.2.3 Higher order rotaxanes
It is more challenging to form [3]- than [2]rotaxanes, for this
reason there are fewer
examples reported in the literature. Several high yielding
[n]rotaxanes, with
n = 3-20, were recently reported by the groups of Stoddart and
Cronin.33
The threads (8 - 10), with n = 0-2, were synthesised in a short
four step synthesis
shown in Scheme 1.3. For the longer threads a protecting group
strategy was
applied.
Scheme 1.3: Thread synthesis in four steps.
The rotaxanes (11 - 13) were synthesised via a clipping process
(Scheme 1.4).
The quaternary ammonium cation templating station proved to be
very effective for
the formation of the polyether macrocycles.
-
Chapter 1
Introduction
12
Scheme 1.4: Templated [n]rotaxane formation with NH2+ as
station.
The aromatic rings of each macrocycle facilitate the formation
of the adjacent
macrocycle via - stacking. The [3]- to [20]rotaxanes (11 - 13)
were formed in
yields ranging from 90 – 93%.
1.2.4 Shuttling
One of the key properties of rotaxanes is the possibility to
shuttle the macrocycle
along the thread. Shuttling can occur when the thread has more
stations than
macrocycles. The stations can be the same or different, as
mentioned in
Section 1.2.1.
-
Chapter 1
Introduction
13
The first molecular shuttle was synthesised by the Stoddart
group in 1991
(Scheme 1.5).24 The rotaxane was obtained via the clipping
method of the
macrocycle around one of the stations. The two stations were
identical and in
deuterated acetone the macrocycle shuttled 500 times per second
between the
two stations.
Scheme 1.5: First molecular shuttle synthesised by Stoddart and
co-workers.
Stoddart reasoned that a switch to a rotaxane with two different
stations would
result in the opportunity to control the shuttling process of
the macrocycle.34 In
1997, Fyfe and Stoddart reported the first switchable
[2]rotaxane incorporating a
benzidine and a biphenol unit.35
The Leigh group has also formed several symmetrical [2]rotaxanes
with two
identical stations (Figure 1.7).36 The amide macrocycle can
shuttle fast between
-
Chapter 1
Introduction
14
the stations in deuterated chloroform, hence only one pair of
signals is observed in
the 1H NMR spectrum. When cooling the sample, the macrocycle
slows down until
it stops, resulting in a splitting of the 1H NMR signals for the
CH2 protons adjacent
to the amide carbonyl on the thread.
Figure 1.7: A symmetrical [2]rotaxane with a tetraamide
macrocycle and a peptide thread.
At room temperature and in deuterated methanol the binding
interactions are
weakened and therefore the rate of shuttling is increased, while
in deuterated
DMSO the shuttling stops completely. In the latter case, the
macrocycle is located
on the hydrophobic part of the thread, enabling the peptide
stations to maximise
their hydrogen bonding interactions with the solvent.
In 2004, Leigh et al. reported the first pH-switchable amide
based molecular
shuttle (Scheme 1.6).37 At room temperature the amide macrocycle
is located over
the succinamide station. When adding a suitable base, such as
potassium
tert-butoxide, the phenol ring gets deprotonated and forms an
anion. This anion is
more attracted to the macrocycle and the shuttling occurs. The
majority of the
macrocycle is then over the phenolate anion. Addition of TFA
results in the
-
Chapter 1
Introduction
15
protonation and reshuttling of the macrocycle to the now more
favoured
succinamide station. A solvent effect was also observed for the
rate of shuttling as
as described in the previous example. This system works in
optimal conditions
with non competing solvents, such as deuterated chloroform.
Scheme 1.6: First pH-switchable amide-based molecular
shuttle.
These are some of the most outstanding examples of shuttling
processes in
rotaxanes, however more examples can be found in the
literature.38-40 The rate of
shuttling is determined by the attractive interactions between
the macrocycle and
the station. When the interactions between the macrocycle and
one station are
strong, the ring does not shuttle and is mainly located on this
particular station.
When the interactions are weak due to repulsive forces, or
solvent effects, or there
are similar strong attractions on the two stations, the
macrocycle can shuttle to
and fro along the thread.
-
Chapter 1
Introduction
16
1.2.5 Rotaxanes used for sensing
Another main application of rotaxanes is the recognition and
binding of ions.41
Beer et al. has recently reported a sulphate-selective binding
and sensing
[3]rotaxane (Scheme 1.7).42
Scheme 1.7: Anion templated synthesis of a [3]rotaxane.
-
Chapter 1
Introduction
17
This rotaxane was synthesised via the anion template clipping
process in 33%
yield. The incorporation of the naphthalene group allows the
rotaxane to exhibit
room temperature fluorescence. The chloride counter-ions were
exchanged by
PF6- anions to enable an easier conversion to other anions. When
replacing just
one of the PF6- anions with one SO4- anion,
the fluorescence was quenched.
However, exchange of the second PF6- anion resulted in an
enhancement of the
emission. The reason for the quenching and enhancement of the
emission lies in
the structural changes that the rotaxane suffers. The first SO4-
anion causes the
rotaxane to bend both macrocycles towards the anion. When the
second PF6-
anion is exchanged by a SO4- anion, each macrocycle captures one
SO4- anion
and the rotaxane returns to its linear geometry.
Catenanes can also be synthesised via the ion template clipping
technique and
these catenanes can sense and bind to various ions.43
1.2.6 Threaded structures with cross-linked macrocycles
Stoddart has also explored other interlocked structures beyond
catenanes and
rotaxanes.44 This new class, of anthropomorphic molecules were
named suitanes,
which consist of two independent compounds: the ‘body’ and the
‘limbs’. The body
could be compared to an unstoppered thread and the limb to a
macrocycle. The
number between suit and ane indicates how many limbs were used
to form the
suitane, e.g. a suit[2]ane has one torso and two limbs and a
suit[3]ane has one
torso and three limbs.
-
Chapter 1
Introduction
18
Scheme 1.8: The three steps for the synthesis of a
suit[2]ane.
Suitanes are usually formed in three steps (Scheme 1.8). In the
first step a
torso-limb complex is formed. The torso (illustrated in blue)
requires at least two
stations (shown in yellow). These stations are capable of
interacting with the limbs,
which are highlighted as red squares. The limb components are
then covalently
linked under thermodynamic control. In the final synthetic step,
the "suit" is
kinetically fixed, which does not allow the components of the
suitane to
disassemble.
In order to synthesise the suit[2]ane 14, three components are
required (Scheme
1.9). For this synthesis, Stoddart and co-workers used a dynamic
covalent
protocol, which they developed in 2004.45
-
Chapter 1
Introduction
19
Scheme 1.9: Components used to form the first suit[2]ane.
Computer calculations suggested that the polyether macrocycles
recognise the
NH2+ stations and form a pseudo[3]rotaxane.† This
pseudo[3]rotaxane can be
captured via imine condensation of p-phenylenediamine and the
two aldehydes,
which are located on top of each macrocycle, to form the
suit[2]ane 14 (Figure
1.8). Even when the imines were not reduced to amines, it was
found that imine 14
was very stable and could be considered a suit[2]ane. To ensure
that the suitane
is in fact an interlocked structure, the sample was heated in
deuterated acetonitrile
for 30 days at 70 °C and no change was observed in the 1H NMR
spectrum.
Suitanes could be considered as a threaded cage-like
molecule.
†Pseudorotaxanes are unstoppered rotaxanes. By adding stoppers
at the ends of the pseudorotaxanes, they become rotaxanes.
-
Chapter 1
Introduction
20
Figure 1.8: Crystal structure of suit[2]ane 14.44
Another method to cross-link two macrocycles was described by
Vögtle et al. in
1996.46,47 An initial sulfonamide [3]rotaxane was synthesised in
4% yield via a
three component clipping process (Scheme 1.10).
Scheme 1.10: Synthesis of a sulfonamide [3]rotaxane.
-
Chapter 1
Introduction
21
This [3]rotaxane can then be cross-linked on the sulfonamide
units of the two
macrocycles using flexible or rigid halide containing chains
(Scheme 1.11).
Scheme 1.11: Cross-linking event to form bonnanes.
This newly formed architectures can be obtained in yields
ranging from 65 - 90%
and are called bonnanes.
1.3 Cage structures
Cage molecules have received great attention over the last
decades, driven by the
surge in ‘bottom-up’ strategies to create small and defined
supramolecular
devices. These devices can then be modified for specific
purposes and as a result
there are many reports of novel syntheses in the literature.
-
Chapter 1
Introduction
22
1.3.1 Metal organic cages
In 1996, Fujita postulated that transition metals could
coordinate compounds to
form definite structures. The coordinated product is therefore
the result of a
templated synthesis and this can be controlled by the structure
and size of the
building blocks.48 Following these principles, Fujita has
described a
self-assembled nanometre-sized organic host framework (Scheme
1.12).49
Scheme 1.12: Self-assembled nanometre-sized organic host
framework synthesised by Fujita.
More recently, Fujita and co-workers showed that these compounds
are able to
function as molecular flasks and help to perform unusual
reactions.50 One of such
unusual reaction is the naphthalene Diels-Alder reaction (Scheme
1.13). These
reactions normally require very harsh conditions (e.g. high
temperature) due to the
lack of reactivity of the naphthalene molecule. However,
quantum-mechanical
calculations showed that this reaction is exothermic. Therefore,
the entropic cost
of the multi-centered reaction is important. The molecular flask
(shown as a circle)
-
Chapter 1
Introduction
23
pre-organises the reactants and greatly reduces the entropic
cost, allowing the
Diels-Alder reaction to occur.
Scheme 1.13: Unusual Diels-Alder reaction promoted by Fujita's
cage.
The reaction profile was changed from a bimolecular pathway to
a
pseudo-intramolecular one.
1.3.2 Boronic ester based cages
Boronic acids are very powerful motifs to assemble molecular and
polymeric
nanostructures.51 The condensation reaction of 1,2- and
1,3-diols with boronic
acids is reversible and the equilibrium is established very
quickly.52
In 2008, Kobayashi and co-workers reported a self-assembled
boronic ester based
cage.53 The tetraboronic acid 15 and the di-catechol linker 16
formed the cage 17
via a condensation reaction in quantitative yield (Scheme 1.14).
The cage 17 is
capable of encapsulating several guest molecules, such as
biphenyl or anthracene
derivatives. This encapsulation process is solvent dependent,
e.g. by removal or
addition of methanol, an on/off switch was observed.
-
Chapter 1
Introduction
24
Scheme 1.14: Synthesis of the boronic ester cage 17, being 16
the four edges of the cage.
Severin et al. recently reported the first example of a
molecular cage with dative
B-N bonds (Scheme 1.15).54 Boronate esters are known to be Lewis
acidic and
therefore the B-atoms are capable of forming dative bonds to
N-donors, such as
pyridines.
Scheme 1.15: Synthesis of the first boronic ester cage with
dative B-N bonds.
-
Chapter 1
Introduction
25
These cage molecules were obtained via multicomponent reactions
of diboronic
acids with catechols and tripyridyl linkers. The cages have six
dative B-N bonds
and the length of the cages can be varied by the choice of the
diboronic acid
component. The cages were not particularly stable, however they
were capable of
encapsulating polyaromatic compounds such as triphenylene.54
1.3.3 Imine cages
Recently, Cooper and co-workers have synthesised the imine cage
20 via
condensation reaction of the trialdehyde 18 and the diamine 19
(Scheme 1.16).55
This particular imine cage is able to incorporate (bind) around
12 to 13 water
molecules. This result is very impressive given that imine bond
formation is
reversible by Nature. Even after 4 hours in refluxing water, the
cage did not
decompose. However, exposure to acidic media led to
decomposition of the cage
since the equilibrium was shifted to the starting aldehyde and
amine.
Scheme 1.16: Synthesis of the imine cage 20.55
-
Chapter 1
Introduction
26
Cage-like molecular structures are interesting for a number of
applications, such
as compound separations,56 catalysis,57-59 sensing,60 and drug
delivery61,62. The
thermo- and hydrolytic stability of the cages is the main
drawback for these
applications.63 This imine cage 20 addresses the hydrolytic
stability issue and
might have opened a new pathway that requires further
investigation for future
applications. A tertiary amine based cage might be able to solve
the thermo- and
hydrolytic problems.
A great number of examples of cage-like structures more or less
closely related to
the areas described in this introduction have been reported.
Further examples
have been reported in the literature and these include MOFs,64
MORFs,65
PMOFs,66 hemicarcerands,67 hydrogen bond capsules68,69 and
COFs70.
-
Chapter 2
Results and Discussion
27
2 Chapter 2: Results and Discussion
2.1 Research aims
As discussed in Chapter 1 many research groups have synthesised
molecular
cages (boxes) via different approaches. The aim of this project
is to synthesise a
novel molecular box via the construction of a [3]rotaxane as key
intermediate.
The first step towards the formation of a molecular box is to
access a thread,
which should be able to form a [3]rotaxane. Thread 21 is a
double fumaric thread
with diphenylethyl secondary amides as stopper groups (Figure
2.1).
Figure 2.1: Double fumaric thread 21.
Its structure is closely related to that of the fumaric thread
used by Leigh and
co-workers in 2001, which allowed them to synthesise a
[2]rotaxane in high yield
(see Section 1.2.2).
In order to build a molecular box, a [3]rotaxane is necessary.
If the desired
[3]rotaxane 22 (Scheme 2.1) can be synthesised, a cross linking
event would be
required to connect the two macrocycles. The nitrogen of the
amide of the
macrocycle or alternative functional groups on the aromatic
rings should allow this
linking process. If functional groups on the aromatic rings are
used, a suitane type
-
Chapter 2
Results and Discussion
28
structure would be formed (see Section 1.2.6). Connecting the
two macrocycles by
using the amide groups would lead to a tertiary amide cage.
Reduction of these
amides to amines, followed by linking would lead to a tertiary
amine cage.
Scheme 2.1: Retrosynthesis for the desired templated [3]rotaxane
22.
When the two macrocycles are linked, a “threaded box” is formed.
A method by
which to liberate the box must be then developed (Scheme
2.2).
-
Chapter 2
Results and Discussion
29
Scheme 2.2: Proposed synthesis of a novel molecular box.
This could be achieved with a cleavable spacer group or by
removing the
stoppers. This would yield a novel molecular box, synthesised by
an unusual
route, which hopefully will enable a novel access to even more
complex structures.
2.2 Attempted syntheses of the double fumaric thread 21
Three main synthetic routes (A, B and C) towards the synthesis
of thread 21 were
identified (Figure 2.2).
Figure 2.2: The three main synthetic routes (A, B and C) towards
the fumaric thread 21.
-
Chapter 2
Results and Discussion
30
Route A involves the synthesis of a stoppered station. Then, two
of these
stoppered stations have to be connected via a spacer group to
from thread 21. In
route B, the two stations are already connected via the spacer
group and a
stoppering event would afford the desired thread 21. Route C
involves the
synthesis of a stoppered half thread, which could be dimerised
to form the desired
fumaric thread 21.
2.2.1 Synthetic route A towards thread 21
In the synthetic route A, the desired fumaric thread 21 could be
formed by reaction
of the known fumaric acid 23 and different diamines in one or
two steps (Scheme
2.3). The fumaric acid 23 could be generated by reaction of the
protected acid
chloride 25 and 2,2-diphenylethylamine, followed by deprotection
of the carboxylic
acid.
Scheme 2.3: Retrosynthetic analysis of route A towards synthesis
of the thread 21.
-
Chapter 2
Results and Discussion
31
Treatment of the acid chloride 26 with 2,2-diphenylethylamine
gave the ester 27 in
94% yield. The ester 27 was hydrolysed to afford the fumaric
thread precursor 23
in quantitative yield. The acid chloride 26 is commercially
available, however it is
expensive and it was synthesised in quantitative yield in one
step from mono-ethyl
fumarate and thionyl chloride under reflux (Scheme 2.4).
Scheme 2.4: Synthesis of the fumaric thread precursor 23.
Once the fumaric thread precursor 23 was synthesised several
amide coupling
reactions were investigated. Since the fumaric thread precursor
23 is a carboxylic
acid, it needs to be activated prior the reaction with the
diamine.
Computational calculations on the fumaric thread 21 with
1,3-diaminopropane as
spacer group showed that four carbon atoms are sufficient to
form four linkages
between the two macrocycles to build a “threaded box” (Figure
2.3). These
calculations‡ were performed using ArgusLab71 and optimised from
the crystal
structure of the [2]rotaxane 2 (see Section 1.2.2) synthesised
by Leigh. The results
from this calculation were then exported to Mercury72 to be
visualised.
‡ Molecular mechanics universal force field 1000 steps.
-
Chapter 2
Results and Discussion
32
Figure 2.3: Threaded box model created via computational
calculations.
Amide coupling reactions are very important in organic
chemistry, and therefore
they are well established. However, these reactions are not
trivial and there are
many different coupling reagents commercially available.73
Unfortunately, there is
still not a universal coupling reagent and if the coupling does
not occur, different
coupling reagents need to be sought.
The fumaric thread 6 was used as a model for the activation of
the thread
precursor 23 via an amide coupling. The fumaric thread 6 was
therefore
synthesised using N,N′-dicyclohexylcarbodiimide (DCC) as
coupling reagent
(Scheme 2.5).
-
Chapter 2
Results and Discussion
33
Scheme 2.5: DCC coupling of the thread precursor 23 with
2,2-diphenylethylamine.
Coupling of the fumaric thread precursor 23 with
2,2-diphenylethylamine was
successful, however it was low yielding (37%).
Despite the low yield of this test reaction, several diamines
(Figure 2.4) were
tested in a coupling reaction with the thread precursor 23
(Table 2.1).
Figure 2.4: Different diamines for use as spacer groups.
None of the conditions tested were suitable for the synthesis of
the double fumaric
thread 21.
All reactions with ethylenediamine, 1,3-propyldiamine and
1,4-butyldiamine
resulted in a complex mixture of side products or decomposition
of the starting
materials.
-
Chapter 2
Results and Discussion
34
Table 2.1: Different coupling reaction conditions studied.
Entry Reagent Additive Solvent Time Temp. Result
1 DCC - THF 48 h RT complex mixture
2 DCC HOBt THF 48 h RT complex mixture
3 HATU Hünigs base CH3CN 24 h RT complex mixture
4 HATU Hünigs base DCM 24 h RT complex mixture
5 EDC - MeOH 16 h RT 50% yield a)
6 EDC HOBt DCM 16 h 0 °C to RT complex mixture
7 NMM / IBCF - THF 30 min -15 °C 27% yield a)
8 SOCl2 - neat 3 h 80 °C - b)
9 SOCl2 - DCM 3 h 80 °C - b)
10 (COCl)2 DMF DCM 30 min 70 °C - b)
11 DCC - DCM 1 h 150 °C complex mixture
a) Reaction occurred only with p-phenylendiamine, however only
the mono reacted product 28 was detected (Scheme 2.6). b)
Decomposition of the starting materials occurred.
These reactions were followed by Matrix-Assisted Laser
Desorption Ionisation
(MALDI) and did not show the expected mass peaks. Figure 2.5
shows the
general principle behind MALDI-ToF mass spectrometry. A sample
of analyte
(shown in red) dispersed within a highly absorbing matrix is
irradiated by a UV
laser beam. The most commonly used UV lasers for MALDI
spectrometry are
nitrogen lasers (337 nm) and Nd:YAG lasers (355 nm and 266 nm).
During the
irradiation, a thin upper layer of matrix gets ablated. This
method of energy
transfer is very mild and excessive energy, which could lead to
decomposition, is
not required. The matrix must absorb the UV laser light
extensively and
2,5-dihydroxy benzoic acid, also called gentisic acid, is used
for this purpose.
http://en.wikipedia.org/wiki/Nd:YAG_lasershttp://en.wikipedia.org/wiki/Gentisic_acid
-
Chapter 2
Results and Discussion
35
Gentisic acid can work in a mass range from 500 to 3,000 dalton
(Da). Below
500 Da, signals corresponding to the matrix can be observed.
Figure 2.5: Schematic representation of MALDI spectrometer.
The mono reacted product 28 was obtained in two different
reaction conditions
when using p-phenylenediamine as diamine. Coupling of 23 with
1-ethyl-3-(3-
dimethylaminopropyl)carbodiimide (EDC) resulted in the mono
reacted product 28
in 50% yield (Scheme 2.6). N-methylmorpholine (NMM) and isobutyl
chloroformate
(IBCF), which forms a mixed anhydride of the carboxylic acid,
gave 28 in 27%
yield.
Scheme 2.6: EDC mediated amide coupling to the mono reacted
product 28.
Unfortunately, it was not possible to transform the mono reacted
product 28 into
the desired double fumaric thread 29 despite screening a number
of peptide
-
Chapter 2
Results and Discussion
36
coupling reagents (Scheme 2.7). Instead, only the starting
material 28 was
recovered.
Scheme 2.7: Attempted reactions towards the synthesis of the
double fumaric thread 29.
After the failure of route A to provide the desired thread 21,
route B was
investigated, whereby the diamine component is introduced in the
first step.
2.2.2 Synthetic route B towards thread 21
In synthetic route B, the proposed retrosynthetic analysis
involves the reaction of
the diamine 30 with the protected fumaric acid chloride 25 to
yield the diamide 31,
which could be deprotected and coupled with
2,2-diphenylethylamine to afford the
fumaric thread 21 (Scheme 2.8).
-
Chapter 2
Results and Discussion
37
Scheme 2.8: Retrosynthetic analysis of route B towards the
synthesis of the thread 21.
The reactions between ethyl fumaroyl chloride and different
diamines were
optimised easily by varying the amount of solvent, reaction time
and amount of
base. Optimisation allowed access to the diamides (32 - 34) in
yields ranging from
42% to 65% (Scheme 2.9).
Scheme 2.9: Synthesis of diamide thread precursors.
-
Chapter 2
Results and Discussion
38
These diamides could potentially form pseudorotaxanes (see
Section 2.6).
Considering the moderate yields for the reactions of the diamide
thread
precursors, an alternative synthetic approach was
investigated.
Katrizky et al. have reported the formation of amides mediated
by NEt3 starting
from benzotriazole and an amine.74 As shown in Scheme 2.10 the
synthesis of the
benzotriazole 35 was high yielding. Unfortunately, the reaction
with ethylendiamine
proceeded in only 33% yield.
Scheme 2.10: Benzotriazole synthetic route to diamide thread
precursor 32.
Saponification of the ester groups of the diamide thread
precursor 33 was
performed in quantitative yield using sodium hydroxide in
methanol (Scheme
2.11). The resulting diacid 36 was used in several coupling
reactions with
2,2-diphenylethylamine.
-
Chapter 2
Results and Discussion
39
Scheme 2.11: Saponification of the diamide thread precursor
33.
Amide coupling reactions of 36 with EDC or DCC as the coupling
reagent did not
form the desired product and the starting material was
recovered. The major
reason for this result may be the lack of solubility in organic
media of the diacid 36.
Therefore an activation of the diacid as the acid chloride
intermediate followed by
reaction with 2,2-diphenylethylamine was tested. Unfortunately,
all attempted
approaches were unsuccessful and resulted in recovery or
decomposition of the
starting material.
Due to the insolubility and unreactivity of the diacid 36, an
alternative approach to
the double fumaric thread 21 was investigated.
The direct synthesis of an amide starting from an ester and an
amine is well
known. This type of amide coupling is normally catalysed by
Zr(OtBu)4 with
1-hydroxy-7-azabenzotriazole (HOAt) and can be very potent, as
described by
Porco (Scheme 2.12).75
Scheme 2.12: Zr-mediated amide coupling.
-
Chapter 2
Results and Discussion
40
These reaction conditions were used to synthesise the known
fumaric thread 6;
which was isolated in 60% yield (Scheme 2.13).
Scheme 2.13: Zr-mediated amide coupling to synthesise thread
6.
Unfortunately, under analogous conditions conversion of the
ester 33 to the amide
37 was not observed (Scheme 2.14). The double amide thread 33
was not soluble
in any solvent previously described for this reaction. The
thread 33 is readily
soluble in methanol, however methanol is not a suitable solvent
for this reaction.
Scheme 2.14: Attempted Zr-mediated amide coupling.
In view of these disappointing results, an alternative synthetic
route was
investigated (Scheme 2.15). The reaction of the ester 22 with
ethylenediamine
was carried out in different solvents using various Lewis acids.
However in no
instance was the desired product obtained. A possible
explanation for this lack of
-
Chapter 2
Results and Discussion
41
reactivity could be binding of ethylenediamine to Zr,
effectively poisoning the
catalyst.
Scheme 2.15: Attempted Zr-mediated amide coupling using
ethylenediamine and ester 22.
As alternatives to Zr(OtBu)4, aluminium chloride and zirconium
chloride were used
in stoichiometric quantities. These catalysts did not show any
reactivity either.
It is possible that the aluminium chloride and zirconium
chloride are not strong
enough to promote this reaction. Unfortunately, Zr(OtBu)4 seems
to be very
powerful and produces numerous products. MALDI analysis did not
show the
desired mass peaks of either single or double coupling
products.
2.2.3 Synthetic route C towards thread 21
Route C does not involve the use of diamines as spacer groups
(Scheme 2.16).
The key intermediate is the amide 38 bearing a functional group
suitable for
dimerisation to afford the desired thread 21, e.g. via
Grubbs-metathesis, or ether
-
Chapter 2
Results and Discussion
42
or anhydride formation. The amide 38 could be formed by reaction
between the
fumaric acid 23 and an amine with an appropriate functional
group.
Scheme 2.16: Retrosynthetic analysis of route C towards the
synthesis of the thread 21.
As previously discussed, several attempts to couple the thread
precursor 23 with
diamines proved unsuccessful. However, a mono reacted amide
coupling
mediated by EDC was detected (see Scheme 2.6). Therefore, EDC as
coupling
agent was tested for the coupling of acid 23 and allylamine
(Scheme 2.17).
Scheme 2.17: EDC mediated amide coupling between the thread
precursor 23 with allylamine.
-
Chapter 2
Results and Discussion
43
The desired product was observed however its purity and yield
were low. For this
reason several attempts to optimise this reaction were carried
out (Table 2.2).
Table 2.2: Different coupling reaction conditions studied for
the thread precursor 23 with allylamine.
Entry Reagent Additive Solvent Time Temp. Result
1 DCC - THF 48 h RT 7%
2 EDC - MeOH 18 h RT 18%
3 NMM/IBCF - THF 30 min -15 °C 3%
However, the EDC mediated coupling reaction gave the best
results. Since the
thread precursor 39 was synthesised in only 18% yield and the
product was not
completely pure, this approach was not investigated any further.
Nevertheless, a
possible reaction could be dimerisation by Grubbs-metathesis.
The
Grubbs-metathesis should lead to an E / Z mixture, which might
be difficult to
purify.
A Zr(OtBu)4 mediated coupling reaction was also tested (Scheme
2.18). However,
only the starting ester 27 was recovered. This might be due to
reaction or
coordination of allylamine with to the Zr. Therefore, no other
Lewis acid was
investigated.
Scheme 2.18: Attempted Zr-mediated amide coupling.
-
Chapter 2
Results and Discussion
44
The desired thread 21 has a diphenylethane motif as the
secondary amide
stopper, however this particular stopper group did not prove
easily amenable for
the proposed synthesis of the thread 21 (Sections 2.2.1 to
2.2.3). Therefore, an
alternative thread system with similar templating-effect was
investigated.
2.3 Synthesis of azo rotaxanes
In July 2010, Berná et al.76 reported an azodicarboxamide thread
system (Scheme
2.19) with a similar templating-effect to the fumaric threads
described by Leigh and
co-workers (see Section 1.2.2).
Scheme 2.19: Synthesis of the azodicarboxamide [2]rotaxane 41 by
Berná.76
To gain a fuller understanding of this new thread system the
previously reported
thread 40, as well as its building blocks, and the [2]rotaxane
41 were synthesised.
The yields obtained were comparable to those reported and were
easily
reproducible (Scheme 2.20).
-
Chapter 2
Results and Discussion
45
Scheme 2.20: Synthesis of the known azodicarboxamide [2]rotaxane
41.
In the first step, phenylchloroformate was reacted with
hydrazine to form the
hydrazophenylcarbamate 42 in 91% yield. Phenylcarbamates are not
very
reactive, however they can be activated with heat and the
addition of a base (e.g.
triethylamine).77 The phenylcarbamate 42 reacted with
dibenzylamine under basic
conditions in refluxing CHCl3 to form the hydrazoamide 43.
Oxidation with
N-bromosuccinimide (NBS) afforded the azo thread 40. In a
clipping process the
azo [2]rotaxane 41 was formed using p-xylylenediamine and
isophthaloyl
dichloride. The yield was significantly lower than that reported
by Leigh and
co-workers for their [2]rotaxane, which suggests that the
templating station is not
as potent as the fumaric one. Despite the lower yield for the
clipping process, the
-
Chapter 2
Results and Discussion
46
efficient synthesis of the azo [2]rotaxane 41 might to enable a
new route towards
the synthesis of the desired [3]rotaxane.
Following an analogous pathway to that described for synthetic
route A in
Section 2.2.1, the azo thread 46 could be obtained via oxidation
of the hydrazo
thread 45 using NBS (Scheme 2.21). The desired hydrazo thread 45
could be
formed by reaction of the hydrazophenylester 44 with various
diamines. The
hydrazophenylester 44 can be generated from the reaction between
the
diphenylester 42 and dibenzylamine.
Scheme 2.21: Retrosynthesis of azo thread 46.
Berná et al. described the synthesis of the building block 44 in
74% yield, which
was used to form a two station [2]rotaxane with the ability to
shuttle its macrocycle.
-
Chapter 2
Results and Discussion
47
This building block was synthesised in 81% yield and purified
via recrystallisation
instead of by column chromatography as Berná described.
Scheme 2.22: Synthesis of the hydrazophenylester 44.
By conducting the reaction at room temperature and decreasing
the number of
equivalents of dibenzylamine used, the formation of the
mono-reacted
hydrazophenylester 44 is favoured over the di-reacted azo
dicarboxamide 43.
2.3.1 Synthesis of azo rotaxanes with a three carbon spacer
unit
The initial studies on the double fumaric thread system
indicated that the use of
1,3-diaminopropane provided the highest yields compared to other
diamines
(see Section 2.2.2). Therefore, hydrazophenylester 44 was
coupled with
1,3-diaminopropane in refluxing chloroform using NEt3 as base
(Scheme 2.23).
The resulting reaction mixture was challenging to purify. Column
chromatography
resulted in crystallisation of the new hydrazo thread 47 on the
column and only
38% of the product was isolated. An attempt to recrystallise the
crude hydrazo
thread 47 proved unsuccessful. Therefore, the moderate yield of
38% after
chromatography was deemed acceptable.
-
Chapter 2
Results and Discussion
48
Scheme 2.23: Synthesis of the hydrazo thread 47.
The hydrazo thread was then oxidised to the corresponding azo
thread 48 using
NBS (Scheme 2.24).
Scheme 2.24: Oxidation of the hydrazo thread 47 to the
corresponding azo thread 48.
It is noteworthy that in the process of changing the double
fumaric thread
(described in Section 2.2) to the azo thread, three important
modifications to the
synthetic route and to the overall thread structure have been
introduced:
the templating station has been altered from fumaric to
azo-dicarboxylate;
-
Chapter 2
Results and Discussion
49
the activated species to couple the station with the spacer unit
has been
changed from an activated acid to a phenylester;
the amines that generate the stopper groups have been modified
from
primary amines to secondary amines.
The azo thread 48 was then used in a rotaxane formation reaction
via the clipping
method using p-xylylenediamine and isophthaloyl dichloride.
MALDI analysis of the crude sample suggested the formation of a
[2]rotaxane with
a mass of 1192 Da and a [3]rotaxane with a mass of 1726 Da
(Figure 2.6).
Figure 2.6: MALDI spectrum of rotaxane reaction using azo thread
48.
1088
.9
1192
.9
1887
.7
556.
5
1726
.61621
.4
0
200
400
600
Inte
ns. [a
.u.]
600 800 1000 1200 1400 1600 1800 2000m/z
-
Chapter 2
Results and Discussion
50
Both of the mass peaks suggested that complete reduction of the
azo stations had
occured during the clipping process.
When attempting to purify the crude sample the major product of
this reaction was
found to be the unwanted catenane described by Leigh and
co-workers,78
accompanied by three different rotaxanes, the [2]rotaxanes 49
and 50 and the
[3]rotaxane 51 (Scheme 2.25).
Scheme 2.25: Synthesis of three different azo rotaxanes with a
three carbon spacer unit.
-
Chapter 2
Results and Discussion
51
High Pressure Liquid Chromatography (HPLC) proved to be very
useful for the
identification and isolation of the diazo [2]rotaxane 50 and the
partially reduced
[2]rotaxane 49 (Figure 2.7).
A gradient method using a ramp from 100% water to 100% methanol
over a period
of 1 hour was used on a HPLC reverse phase column at a flow rate
of 1 mL/min.
This initial chromatogram helped to determine the ratio between
the [2]- and
[3]rotaxanes in the crude mixture. The presence of [2]rotaxanes
was in an excess
of 10:1 with respect to the desired [3]rotaxane. At this stage
it was not obvious that
two different [2]rotaxanes were formed.
There are several possible explanations for this product
distribution. Assuming that
the [2]rotaxane is formed first, the macrocycle could shuttle
rapidly between the
two stations and diminish the formation of the [3]rotaxane 51.
Another possibility is
that the two macrocycles are too close together such that the
first formed
macrocycle blocks the second one from assembling at the other
station.
-
Chapter 2
Results and Discussion
52
Figure 2.7: Analytical HPLC chromatogram of the rotaxane
reaction using the azo thread 48. The first peak (39.87 min)
corresponds to the reduced [2]rotaxane 49, the peak at 40.14 min
corresponds to the diazo [2]rotaxane 50 and the third highlighted
peak at
40.64 min corresponds to the [3]rotaxane 51.
-
Chapter 2
Results and Discussion
53
2.3.2 Synthesis of azo rotaxanes with a four carbon spacer
unit
In order to further investigate these hypotheses, an azo thread
with a four carbon
spacer unit was synthesised. Initially, the hydrazophenylester
44 was coupled with
1,4-diaminobutane in refluxing chloroform using NEt3 as base
(Scheme 2.26).
Scheme 2.26: Synthesis of the azo thread 53.
The newly synthesised hydrazo thread 52 presented similar
purification issues to
those observed for the hydrazo thread 47. Therefore, the thread
52 could only be
purified via column chromatography in 32% yield. Oxidation of 52
using NBS
afforded the azo thread 53 in 73% yield.
The azo thread 53 was subsequently reacted under the previously
described
clipping conditions and analysed via HPLC and MALDI. The HPLC
chromatogram
showed an increased in the formation of the [3]rotaxane, with a
ratio of 6:1
between the [2]- and [3]rotaxanes.
-
Chapter 2
Results and Discussion
54
Scheme 2.27: Synthesis of three different azo rotaxanes with a
four carbon spacer unit
However, only the two different [2]rotaxanes could be isolated
using HPLC. When
trying to isolate these [2]rotaxanes via flash column
chromatography an interesting
phenomenon was observed. The [2]rotaxane 54 and the free
macrocycle co-eluted
with nearly a perfect 1:1 ratio. Consequently, the 1H NMR
spectrum suggested the
isolation of the [3]rotaxane 56, which contradicted the results
observed by ES+
and MALDI mass spectrometries.
-
Chapter 2
Results and Discussion
55
Figure 2.8: 1H NMR spectrum of [2]rotaxane 54 and free
macrocycle.
Figure 2.8 shows the aromatic region of the 1H NMR spectrum of
the purified
material. A doubling of the signals of the distinctive A, B and
C protons is
observed. The protons with a star correspond to the free
macrocycle and the ones
without, to the [2]rotaxane 54. The integration is nearly 1:1,
which suggested the
isolation of the desired [3]rotaxane. However, the splitting of
the signals for the A,
B and C protons into two pair of sets contradicts the existence
of the [3]rotaxane.
In order to elucidate whether some sort of uncommon conformer
issue was
A A* C C*
B* B
-
Chapter 2
Results and Discussion
56
observed, saturation transfer experiments (GOSY) and variable
temperature NMR
studies (Figure 2.9) were performed.
Figure 2.9: VT-NMR spectra of the 1:1 mixture of [2]rotaxane 54
and free macrocycle.
At lower temperature only the signals of the [2]rotaxane 54 are
sharp and definite,
while the signals of the free macrocycle are broad. The two sets
of signals did not
convert into one set either, which strongly implies the
existence of two different
compounds.
The existence of the two species was further proven by injecting
the NMR sample
into a HPLC column to separate the two compounds. The 1H NMR
spectra of the
+30 ºC
+10 ºC
-10 ºC
-30 ºC
-50 ºC
-70 ºC
-
Chapter 2
Results and Discussion
57
HPLC purified [2]rotaxane 54 showed the expected A, B and C
protons and the
lack of the macrocyclic A*, B* and C* protons.
2.3.3 Synthesis of azo rotaxanes with a five carbon spacer
unit
Since a significant increase in the formation of [3]rotaxane was
observed when
increasing the spacer length (from three to four carbons), a
five-carbon linker was
investigated next. The hydrazophenylester 44 was coupled
with
1,5-pentanediamine in refluxing chloroform using NEt3 as base
(Scheme 2.28).
The hydrazo thread 57 was isolated in moderate 29% yield as it
presented the
same behaviour observed for threads 47 and 52. The hydrazo
thread 57 was
oxidised to the azo thread 58 in 88% yield.
Scheme 2.28: Synthesis of the azo thread 58.
-
Chapter 2
Results and Discussion
58
The azo thread 58 was used to synthesise the desired [3]rotaxane
61, although in
a very low yield (3%). Unsurprisingly, the azo [2]rotaxane 59
and the hydrazo
[2]rotaxane 60 were also observed and these were isolated.
Scheme 2.29: Synthesis of three different azo rotaxanes with a
five carbon spacer unit.
Isolation and characterisation of the desired [3]rotaxane 61 was
realised via an
isocratic HPLC method using a mixture of methanol and water (75
: 25).
Significant quantities of the azo 59 and hydrazo 60 [2]rotaxanes
were also isolated
and fully characterised. Figure 2.10 show the resulting HPLC
chromatogram.
-
Chapter 2
Results and Discussion
59
Figure 2.10: Analytical HPLC chromatogram of the rotaxane
reaction using the azo thread 58. Hydrazo [2]rotaxane 60 at 17.53
min, azo [2]rotaxane 59 at 20.48 min and
[3]rotaxane 61 at 40.17 min.
Attempts to increase the yield of the desired [3]rotaxane 61
were unsuccessful.
Neither slower addition nor changing the monomer concentration
led to an
increase in the formation of the [3]rotaxane 61.
-
Chapter 2
Results and Discussion
60
Figure 2.11: 1H NMR spectra of the three different azo rotaxanes
with a five carbon spacer unit. The hydrazo [2]rotaxane 60 in blue,
the azo [2]rotaxane 59 in green and the
azo [3]rotaxane 61 in red.
Figure 2.11 shows the 1H NMR spectra of the three different azo
rotaxanes with a
five carbon spacer unit. In the aromatic region, the A, B and C
protons were clearly
identified. The aromatic signals of the stopper groups are less
defined for the azo
and hydrazo [2]rotaxanes. The benzylic region is quite complex
and it is expanded
in Figure 2.12. The benzylic protons belong to the two benzylic
groups of the
stoppers and the four benzylic units of the macrocycle.
A B
C
-
Chapter 2
Results and Discussion
61
Figure 2.12: Benzylic region of the 1H NMR spectra of the three
azo rotaxanes with a five carbon spacer unit. The hydrazo
[2]rotaxane 60 in blue, the azo [2]rotaxane 59 in green
and the azo [3]rotaxane 61 in red.
The 1H NMR spectrum for the partially reduced hydrazo
[2]rotaxane 60 (shown in
blue) is rather complex, since the macrocycle is mainly located
over the azo
station and therefore the [2]rotaxane 60 is unsymmetrical.
Consequently, separate
benzylic signals are observed for the stoppers on the azo
station, hydrazo station
and the macrocycle. The aliphatic signals of the [2]rotaxane 60
are also split
indicating that the shuttling is stopped. The 1H NMR spectrum of
the azo
[2]rotaxane 59 (shown in green) presents just three signals in
the benzylic region.
This is due to the shuttling process between the two stations.
Since both stations
are identical, a simpler 1H NMR spectrum is observed. The
[3]rotaxane 61
(1H NMR spectrum shown in red) has two fully occupied stations
and cannot
shuttle, therefore five signals are detected. These observations
are in agreement
with those recently reported by Berná and co-workers.79
hydrazo [2]rotaxane 60
azo [2]rotaxane 59
azo [3]rotaxane 61
-
Chapter 2
Results and Discussion
62
Berná has reported that only the azo stations and not their
reduced counter parts
are able to form a rotaxane in good yield.76 Since a 12% yield
of unwanted
[2]rotaxanes was obtained and given that the oxidised
[2]rotaxane 59 might form
the desired [3]rotaxane 61, it was investigated whether the
reduced [2]rotaxane 60
could be re-oxidised to the [2]rotaxane 59. After the HPLC
purification, the
reduced hydrazo [2]rotaxane 60 could not be completely purified.
Therefore, the
obtained 4:1 (reduced : oxidised) mixture was oxidised to
completion using NBS
(Scheme 2.30).
Scheme 2.30: Oxidation of the [2]rotaxane 60.
A second clipping reaction was then performed on the azo
[2]rotaxane 59 and the
[3]rotaxane 61 was successfully isolated in 22% yield (Scheme
2.31). Not
surprisingly, spontaneous reduction of one of the azo stations
on the [2]rotaxane
-
Chapter 2
Results and Discussion
63
59 to yield 60 was also observed. This is the main drawback of
this approach as
only 13% combined yield of the starting and reduced [2]rotaxanes
could be
obtained.
Scheme 2.31: Clipping reaction starting from [2]rotaxane 59.
Colourless crystals suitable for X-ray diffraction were obtained
by slow evaporation
of a dichloromethane : methanol (1:1) solution of [3]rotaxane
61. The X-ray crystal
structure confirms the formation of a [3]rotaxane as shown in
Figure 2.13, where
the macrocycles are highlighted in blue and the thread in
orange.
-
Chapter 2
Results and Discussion
64
Figure 2.13: Crytal structure of [3]rotaxane 61. Solvent
molecules have been omitted for clarity.
The macrocycles have a twisted boat-like structure. As
previously discussed
(see Section 1.2.2), this conformation suggests that the
critical cyclic syn syn
intermediate around the thread is not effectively formed, which
is also represented
in the low yield obtained. The macrocycles are twisted with
respect to one another
by 38° and they are located at 11.6 Å from each other. Hydrogen
bonds (shown in
purple) are observed between the thread and the macrocycles. The
bond length
distance between donor and acceptor ranges from 2.834(10) to
3.279(9) Å. The
macrocycle on the right is twisted and one carbonyl group can
hydrogen bond to
one NH group of the spacer unit on the thread.
-
Chapter 2
Results and Discussion
65
2.3.4 Spontaneous reduction of azo systems
In order to further investigate the spontaneous reduction
observed on the azo
rotaxanes, a series of experiments were undertaken on the simple
azo thread 40.
This thread system can be readily synthesised in large
quantities in three steps,
providing access to a simple model system on which the
spontaneous reduction
reaction can be studied.
Scheme 2.32: Benchmark reaction for the spontaneous
reduction.
Several reagents to reduce the azo thread 40 to its hydrazo
analogue 43 (Scheme
2.32) were tested and these are shown in Table 2.3.
Table 2.3: Reaction conditions studied for the reduction of the
azo compound 40.
Entry Reagent Amount Time Conversion
1 NEt3 48 eq. 8 h 18%
2 HCl 48 eq. 8 h 1%
3 NEt3 / HCl 48 eq. 8 h 2%
4 DMAP 500 eq. 6 d -
5 NEt3 500 eq. 6 d quant.
When 48 equivalents of NEt3 were used over a period of eight
hours, the same
conditions used in the clipping process, a conversion of 18%
from the azo to the
-
Chapter 2
Results and Discussion
66
hydrazo compound was observed. On the other hand, if HCl was
used a
conversion of only 1% was detected, although this may be due to
a different
reaction pathway. A mixture of NEt3 and HCl yielded 2%
conversion. This result
suggests that the rates of reduction for NEt3 and HCl are
different since a higher
yield was observed when the combination of both reagents was
used. 500
equivalents of NEt3 led to full conversion of the azo compound
40 to the reduced
hydrazo analogue after six days at RT. DMAP did not give any
conversion.
These reactions were monitored by 1H NMR spectroscopy and the
percentage of
conversion was calculated according to the integration of the
different peaks
(Figure 2.14). The benzylic region is significantly different in
the azo and hydrazo
compounds.
Figure 2.14: 1H NMR spectra from the azo thread 40 (in blue),
the hydrazo thread 43 (in green) and the reaction mixture of the
azo thread 40 and 48 eq. of NEt3 (in red).
-
Chapter 2
Results and Discussion
67
The 1H NMR spectrum shown in blue corresponds to the azo thread
40 and
presents two benzylic signals, since 40 is conformationally
locked due to amide
rotamer restriction. However, in the hydrazo thread 43 (1H NMR
spectrum shown
in green) these benzylic protons coalesced at RT and a singlet
is observed. In the
spectrum shown in red, which corresponds to the reaction mixture
of the azo
compound 40 with 48 equivalents of NEt3, a new peak at 4.46 ppm
corresponding
to the hydrazo compound is observed.
As mentioned above, there are several possible mechanisms for
this spontaneous
reduction process. However, to date, it has not been possible to
clearly elucidate
it. The most feasible reaction pathway is depicted in Scheme
2.33.
Scheme 2.33: Proposed mechanism for the formation of the hydrazo
compound 43.
In the first step, NEt3 might add to the azo compound in the
same way that PPh3
does in a Mitsunobu reaction, and the zwitterionic intermediate
I might be formed.
An -hydrogen abstraction, via a five-membered ring transition
state, followed by
-
Chapter 2
Results and Discussion
68
elimination might lead to the negatively charged hydrazo species
II and the
positively charged iminium intermediate III. Hydrolysis might
yield the hydrazo
compound 43, diethylamine and acetaldehyde. This proposed
mechanism could
be considered as an addition / elimination / hydrolysis
sequence. However, the
term reduction seems appropriate in order to compare the azo and
hydrazo
compounds.
2.4 Synthesis of fumaric rotaxanes
An alternative approach intending to overcome the spontaneous
reduction, the low
yield for the formation of the desired [3]rotaxanes, and the
fact that the two
macrocycles are quite separated from each other, was explored.
The idea that a
change of the stopper group could lead to a double fumaric
thread was explored.
2.4.1 Synthesis of a fumaric [2]rotaxane with dibenzylamide
stopper groups
As a starting point, the successful dibenzyl motif used on the
azo rotaxanes was
chosen as the stopper group for the new fumaric threads. The
thread 62 was
synthesised in order to test the effect of using dibenzylamine
as a stopper group in
the formation of a simple thread (Scheme 2.34).
Scheme 2.34: Synthesis of the fumaric thread 62.
-
Chapter 2
Results and Discussion
69
The thread 62 was then reacted under the previously used
clipping reaction
conditions for rotaxane formation and the [2]rotaxane 63 could
be purified via
column chromatography and was isolated in 84% yield (Scheme
2.35).
Scheme 2.35: Synthesis of the [2]rotaxane 63.
In the aromatic region of the [2]rotaxane 63 the A, B, C and D
protons of the
macrocycle were clearly identified.
Figure 2.15: 1H NMR of thread 62 (in blue) and [2]rotaxane 63
(in red).
D
CH
A B
C
-
Chapter 2
Results and Discussion
70
The olefinic protons (highlighted in green) of the thread 62
shifted upfield (from
7.61 to 6.04 ppm) upon formation of the [2]rotaxane 63. This
shielding effect is
well documented and observed in several supramolecular systems,
e.g. the
fumaramides reported by Leigh et al..80 This shielding effect of
the olefinic protons
was observed in all of the following fumaramide based rotaxanes
synthesised.
Colourless crystals suitable for X-ray diffraction were obtained
by slow evaporation
of a chloroform : methanol (10 : 1) solution of [2]rotaxane
63.
Figure 2.16: Crystal structure of [2]rotaxane 63. Solvent
molecules have been omitted for clarity.
The crystal structure confirms the formation of the desired
fumaric [2]rotaxane with
hydrogen bonds between the two O-atoms of the station and the NH
of the
-
Chapter 2
Results and Discussion
71
macrocycle being observed (Figure 2.16). The bond length
distance between
donor and acceptor ranges from 2.9772(18) to 3.2685(19) Å. The
macrocycle is
quite packed in between the two stopper groups, leaving little
space between
them. For this reason, the high yield formation of [2]rotaxane
63 is quite surprising.
On the other hand, the macrocycle has a chair-like structure
indicating that the
cyclic syn syn intermediate around the thread is effectively
formed, which results in
a good yield. The centre of the molecule is located on an
inversion centre such
that only half of the [2]rotaxane is crystallographically unique
with a symmetry
relation of -x,-y,-z. In contrast with the twisting observed for
the macrocycles of the
[3]rotaxane 61, the tetraamide macrocycle of this [2]rotaxane is
at a near right
angle with the thread.
2.4.2 Synthesis of a fumaric [3]rotaxane with a two carbon
spacer unit
Encouraged by these promising results, a new synthetic route
based on that
described in Section 2.2.1 (route A) was investigated. Treatment
of the acid
chloride 26 with dibenzylamine gave the ester 64 in 86% yield.
The ester 64 was
then hydrolysed to afford the fumaric thread precursor 65 in
essentially
quantitative yield (99%). As discussed in Section 2.2, the acid
chloride 26 is
commercially available, however it was synthesised in
quantitative yield in one
step from mono-ethyl fumarate using thionyl chloride under
reflux conditions.
-
Chapter 2
Results and Discussion
72
Scheme 2.36: Synthesis of the thread precursor 65.
The thread precursor 65 was transformed into the corresponding
acid chloride
using thionyl chloride under reflux conditions. Treatment of the
acid chloride with
ethylenediamine gave the desired double fumaric thread 66 in 66%
yield (Scheme
2.37). This positive result emphasises that the previously
chosen amide stopper
group (Section 2.2) was not optimal to create the desired double
fumaric thread.
Scheme 2.37: Synthesis of the thread 66.
The newly synthesised double fumaric thread 66 was able to form
the desired
[3]rotaxane 67 in 8% yield (Scheme 2.38). This [3]rotaxane is
very insoluble in any
-
Chapter 2
Results and Discussion
73
organic solvent other than methanol and had to be isolated via
HPLC using a
mixture of methanol and water (75 : 25). The corresponding
[2]rotaxane was also
detected, however it could not be isolated. Many different
purification techniques
such as column chromatography, recrystallisation or different
HPLC methods were
attempted, however these did not lead to bigger amounts of pure
product.
Scheme 2.38: Clipping reaction to form the [3]rotaxane 67.
Thus, by changing the templating station from azo dicarboxamide
to fumaramide
and shortening the spacer unit from five to two carbons, the
yield of the
[3]rotaxane was significantly increased from 3 to 8%.
-
Chapter 2
Results and Discussion
74
2.4.3 Synthesis of a fumaric [3]rotaxane with a three carbon
spacer unit
In order to investigate whether a longer spacer group would
promote a higher
yield, 1,3-diaminopropane was tested as spacer group. The thread
precursor 65
was transformed into the corresponding acid chloride. Treatment
of the acid
chloride with 1,3-diaminopropane gave the desired double fumaric
thread 68 in
42% yield (Scheme 2.39).
Scheme 2.39: Synthesis of double fumaric thread 68.
The newly synthesised double fumaric thread 68 was able to form
the desired
[3]rotaxane 69 in 17% yield (Scheme 2.40). The purification of
this [3]rotaxane was
challenging, as it presents poor solubility in organic solvents.
Column
chromatography gave a mixture of [3]rotaxane and catenane.
Recrystallisation
was also unsuccessful. However, the [3]rotaxane 69 could be
purified and isolated
via HPLC. The corresponding [2]rotaxane was also formed as a
minor product.
-
Chapter 2
Results and Discussion
75
Scheme 2.40: Clipping reaction to from the [3]rotaxane 69.
Colourless crystals suitable for X-ray diffraction were obtained
by slow evaporation
of a dichloromethane : ethanol (1:1) solution of [3]rotaxane 69.
The crystal
structure confirms the formation of a [3]rotaxane (Figure 2.17).
Hydrogen bonds
between the thread and the macrocycles are observed, the bond
length distance
between donor and acceptor being in a range from 2.670(9) to
3.398(5) Å. The
macrocycles are in a half chair / twisted boat conformation
indicating a
semi-efficient cyclic syn syn intermediate around the thread,
which leads to a
moderate yield. The macrocycles are twisted with respect to one
another by 30°
and they are located at 9.2 Å from each other. A cross coupling
linking event
between both macrocycles seems now more likely to occur, since
the two
macrocycles are closer to each other and the angle between them
has been
-
Chapter 2
Results and Discussion
76
reduced. However, the conformation in the solid state might be
different from that
in solution.
Figure 2.17: Crystal structure of [3]rotaxane 69. Solvent
molecules have been omitted for clarity.
Therefore, by increasing the length of the spacer group in the
double fumaric
thread the yield for the formation of the [3]rotaxane has
increased considerably
(from 8% to 17%). Consequently, a three carbon spacer unit seems
a good
compromise between an acceptable yield for the desired
[3]rotaxane and the
distance between the two macrocycles.
-
Chapter 2
Results and Discussion
77
2.4.4 Synthesis of a fumaric [3]rotaxane with a three carbon
spacer unit and
a phenylbenzyl amide stopper
In 2008, Leigh and co-workers81 reported a secondary amide
stopper group using
phenylbenzylamine. This stopper group was also investigated in
this study in order
to assess the effect of different stopper groups upon rotaxane
formation. The yield
for the formation of the ester 70 was improved using the
previously described acid
chloride protocol. The saponification step was carried out by
using the same
conditions described by Leigh et al. to yield the desired acid
precursor 71 in 68%
yield (Scheme 2.41).
Scheme 2.41: Synthesis of the carboxylic acid 71.
The carboxylic acid 71 was transformed into the corresponding
acid chloride using
thionyl chloride. Treatment of the resulting acid chloride with
1,3-diaminopropane
gave the desired double fumaric thread 72 in an acceptable 84%
yield (Scheme
2.42).
-
Chapter 2
Results and Discussion
78
Scheme 2.42: Synthesis of the fumaric thread 72.
This new double fumaric thread 72 was able to form the desired
[3]rotaxane 73 in
a notable 40% yield (Scheme 2.43). The [3]rotaxane 73 could be
purified by
column chromatography and this might be a reason for the higher
yield compared
to its predecessors that needed to be purified by HPLC methods.
The
corresponding [2]rotaxane 74 was also observed and isolated (5%
yield).
-
Chapter 2
Results and Discussion
79
Scheme 2.43: Synthesis of [3]rotaxane 73.
Colourless crystals suitable for X-ray diffraction were obtained
by slow evaporation
of a dichloromethane : ethanol (1:1) solution of pure
[3]rotaxane 73. The crystal
structure confirms the formation of a [3]rotaxane (Figure
2.18).
-
Chapter 2
Results and Discussion
80
Figure 2.18: Crystal structure of [3]rotaxane 73. Solvent
molecules have been omitted for clarity.
The two macrocycles are twisted with respect to one another by
73º, which is a
greater twist compared to that observed for the [3]rotaxane 61
(30º). However, the
distance between both macrocycles remains 9.2 Å. The macrocycles
are in a
perfect chair conformation, which indicates that the cyclic syn
syn intermediate
around the thread is formed effectively and this is represented
by the higher yield.
Hydrogen bonds are observed between the thread and the
macrocycles, with bond
length distances between donor and acceptor ranging from
2.711(4) to 3.127(3) Å.
2.5 Further reactions towards a molecular cage
The next step to obtain the desired box is to link the two
macrocycles together.
Initially, allylation or propargylation of the N-atoms of the
macrocycles were
investigated for the [2]rotaxane 63. However, these reactions
proved ineffective.
-
Chapter 2
Results and Discussion
81
This might result from the direction in which the amide protons
of the macrocycle
point towards the carbonyl groups of the templating stations,
due to the attractive
hydrogen bonding interaction. The high rigidity of the
tetraamide macrocycle can
also limit the reactivity of the amide protons. In contrast, the
methylation reaction
of [2]rotaxane 63 was successful, as confirmed by MALDI
spectrometry. This
could be due to the smaller size of the methyl groups that might
have enough
space to fit in between the macrocycle and the thread.
Scheme 2.44: Methylation of [2]rotaxane 63.
The reactivity of the nitrogen atoms of the amide macrocycle
could be enhanced
by reduction of the amides to amines. This would create a more
flexible
macrocycle and lead to more reactive N-ato